This project is a Python based classification task using the Iris dataset. The objective is to demonstrate data preprocessing, model training, and evaluation of classification models.
- Clone the repository in an empty directory.
- On Windows run
setup\setup.bat, on Unixsetup/setup.shin the terminal. - In the terminal run
jupyter notebookand visit the webpage using the provided link in the terminal.
The Iris dataset is a widely-used dataset in the machine learning community, featuring measurements of four features (sepal length, sepal width, petal length, and petal width) for three species of Iris flowers: setosa, versicolor, and virginica. Libraries I will be using are:
- pandas: For organizing data.
- matplotlib: For visual aid.
- scikit-learn: Contains predefined algorithms for creating models.
- pickle: For saving and loading models.
The project comprises the following components:
-
Importing packages: Packages that will be useful for the whole process of the development of the model are imported. These packages are listed in the overview above with their function.
-
Data preprocessing: Here, the dataset is loaded and preprocessed. Data preprocessing includes checking for missing values, examining data distribution, and encoding categorical features.
-
Data splitting: After feature and target extraction. The dataset is divided into training and testing sets, enabling the training and evaluation of machine learning models.
-
Training models: Three different classification models are employed in this project:
- Random Forest
- Support Vector Machine (SVM)
- k-Nearest Neighbors (KNN)
-
Model Evaluation: The project calculates and displays accuracy scores for each classification model to assess their performance.
-
Model Packaging: The models are saved with their names using the pickle library and stored in the Iris_Models directory. They can be loaded for use at any later time.